View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Virology 413 (2011) 231–243 Contents lists available at ScienceDirect Virology journal homepage: www.elsevier.com/locate/yviro HIV-1 entry inhibition by small-molecule CCR5 antagonists: A combined molecular modeling and mutant study using a high-throughput assay Jean Labrecque a,1, Markus Metz b,1, Gloria Lau a, Marilyn C. Darkes a, Rebecca S.Y. Wong a, David Bogucki b, Bryon Carpenter b, Gang Chen b, Tongshuang Li b, Susan Nan b, Dominique Schols c, Gary J. Bridger b, Simon P. Fricker a, Renato T. Skerlj b,⁎ a Department of Biology, AnorMED Inc. now Genzyme Corporation, 500 Kendall Street, Cambridge, MA 02142, USA b Department of Chemistry, AnorMED Inc. now Genzyme Corporation, 500 Kendall Street, Cambridge, MA 02142, USA c Rega Institute for Medical Research, Katholieke Universiteit Leuven, Minderbroedersstraat 10, B-3000, Leuven, Belgium article info abstract Article history: Based on the attrition rate of CCR5 small molecule antagonists in the clinic the discovery and development of Received 14 January 2011 next generation antagonists with an improved pharmacology and safety profile is necessary. Herein, we Returned to author for revision describe a combined molecular modeling, CCR5-mediated cell fusion, and receptor site-directed mutagenesis 13 February 2011 approach to study the molecular interactions of six structurally diverse compounds (aplaviroc, maraviroc, Accepted 18 February 2011 vicriviroc, TAK-779, SCH-C and a benzyloxycarbonyl-aminopiperidin-1-yl-butane derivative) with CCR5, a Available online 9 March 2011 coreceptor for CCR5-tropic HIV-1 strains. This is the first study using an antifusogenic assay, a model of the Keywords: interaction of the gp120 envelope protein with CCR5. This assay avoids the use of radioactivity and HIV HIV-1 infection assays, and can be used in a high throughput mode. The assay was validated by comparison with CCR5 other established CCR5 assays. Given the hydrophobic nature of the binding pocket several binding models Small molecule antagonists are suggested which could prove useful in the rational drug design of new lead compounds. Aplaviroc © 2011 Elsevier Inc. All rights reserved. Maraviroc Vicriviroc Mutagenesis Antifusogenic assay Homology modeling Binding site characterization The successful treatment of Acquired Immune Deficiency Syn- membrane, which ultimately results in viral load reduction when drome (AIDS) has been due to the introduction of Highly Active administered to humans (Allen et al., 2007). While CCR5 is used as a Antiretroviral therapy (HAART). Two classes of HIV drugs (e.g. coreceptor in the early stages of infection, the CXCR4 coreceptor using proteases and reverse transcriptases) are used in HAART to delay or virus is linked to significant disease progression leading to AIDS. In stop the progression of AIDS. However, because of side effects, contrast to in vitro results, other chemokine receptors have not been development of resistance, and challenges with patient compliance found to play a role as coreceptors for HIV infection under in vivo there is a continuing demand for new generations of viral inhibitors. conditions (Dragic, 2001; Kazmierki et al., 2005). Ideally, other classes of drugs interfering with viral replication by new The important role of these chemokine receptors in the HIV mechanisms could complement existing therapies (Barber, 2004; infection process was discovered in 1998, which led to a new Lusso, 2006; Markovic, 2006; Nadler and Phillips, 2005). classification of the three main HIV variants based on their chemokine One such new class, the entry inhibitors, prevents HIV entry into receptor selectivity: these classifications are CXCR4-tropic (X4), the target cell. In order for HIV to infect the host cell the gp120 viral CCR5-tropic (R5), and dual-tropic (R5/X4) (Berger et al., 1998). envelope protein must first bind to the CD4 receptor, and then to The development of small molecule antagonists of CCR5 to block HIV either one of two chemokine coreceptors, CCR5 or CXCR4, which are entry has been pursued by several pharmaceutical companies and G-protein coupled receptors, members of the seven-transmembrane multiple compounds have been evaluated in the clinic (Faetkenheuer domain family. Antagonists inhibiting or altering the interaction of et al., 2005; Klibanov, 2009; Kromdijk et al., 2010; Lalezari et al., 2005; HIV with these two coreceptors prevent viral fusion with the cell Lenz and Rockstroh, 2010; Palani and Tagat, 2006; Perry, 2010; Sayana and Khanlou, 2009; Shuermann et al., 2007; Strizki et al., 2005; Tilton et al., 2010; Wilkin et al., 2010). These small molecule antagonists have ⁎ Corresponding author. Fax: +1 781 672 5823. E-mail address: [email protected] (R.T. Skerlj). been found to bind within a binding pocket formed by the transmem- 1 Both authors contributed equally. brane helices (Billick et al., 2004; Dragic et al., 2000; Castonguay et al., 0042-6822/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.virol.2011.02.016 232 J. Labrecque et al. / Virology 413 (2011) 231–243 2003; Kondru et al., 2008; Maeda et al., 2006; Nishikawa et al., 2005; Pharmaceuticals (Maeda et al., 2004) and licensed by GlaxoSmithKline, Stupple et al., 2011; Seibert et al., 2006; Tsamis et al., 2003). The fact that was halted after patients in Phase II and III trials experienced liver toxicity these antagonists are small compared to HIV gp120 and that they bind in (Crabb, 2006). Thus, given the high attrition rate of compounds in the a cavity within the receptor makes it unlikely that a direct competition clinic increasing the understanding of how small molecules bind to CCR5 could occur between these inhibitors and gp120. Instead, mechanistic should help in developing structurally diverse antagonists with different studies have indicated that these antagonists act via an allosteric pharmacokinetic, pharmacodynamic and resistance profiles. mechanism, changing the conformation of the receptor and thus In this paper we provide a comprehensive description of the binding preventing the interaction of gp120 with CCR5 (Dragic et al., 2000; of the aforementioned antagonists to CCR5 using an antifusogenic assay, Watson et al., 2005). Though the details of this allosteric mechanism are which mimics the interaction of the viral envelope protein with CCR5. not fully understood, the concept of allosteric inhibition provides the After validating this assay with other in-house data (natural ligand opportunity to develop structurally diverse CCR5 antagonists with binding and HIV-1) and by comparing the effect of selected CCR5 amino different binding interactions potentially inducing different conforma- acid mutations on small molecule inhibition, an in-depth mutant study of tional changes of the receptor which are incapable of interacting with the potential allosteric binding site has been conducted. The conclusions the virus leading to different resistance profiles. To date only one CCR5 of this study are based primarily on our fusion assay data, with additional antagonist has successfully progressed through all the developmental literature information used to refine our small molecule binding hurdles. Pfizer's maraviroc was approved by the FDA for use in treatment hypotheses. This work is in line with other mechanistic studies aimed experienced patients in August 2007 (Perry, 2010; Sayana and Khanlou, at a better understanding of small molecule binding and inhibition of 2009) and following further clinical trials was approved for use in CCR5. Notably, use of the fusion assay provides a physiologically relevant treatment-naïve patients with CCR5-using virus in November 2009 technique for investigating drug/receptor interactions with the merits of (Kromdijk et al., 2010).Otheragentshavebeenlesssuccessful.Asthefirst not requiring active virus, and within a short assay time. Furthermore the of its class, TAK-779 was tested in clinical trials, but failed because of poor binding models provided by this assay, and reported in this paper, can pharmacological and/or toxicological properties as well as lack of oral provide valuable insights for rational drug design. bioavailability (Palani and Tagat, 2006). Schering-Plough has progressed two compounds (SCH-C and vicriviroc) into the clinic. However, Phase I Results studiesofSCH-CwerediscontinuedbecauseofprolongationoftheQT interval and Phase II studies involving treatment-naïve patients with Validation of a huCCR5/gp120 fusion assay as a technique for vicriviroc were stopped because of the increased likelihood of viral load interrogating compound/receptor interactions using receptor reboundcomparedtoagroupofpatientsusingstandardtherapy.In site-directed mutagenesis addition based on results of two Phase III trials of vicriviroc in treatment- experienced HIV-positive patients, vicriviroc did not meet its primary The amino acid sequence of CCR5 is shown as snake plot in Fig. 1 efficacy endpoint therefore, Merck will not submit an NDA for vicriviroc together with 34 single-site mutants to encompass all possibilities for (Merck, 2010). Development of aplaviroc, originally developed by Ono small molecule binding sites (Table 1). Fig. 1. Snake plot representation of CCR5; the mutated residues tested in the current study are highlighted and important residues are annotated. J. Labrecque et al. / Virology 413 (2011) 231–243 233 Table 1 inhibitor binding as reported in the literature: E283A, W86A, Y37A, Replicate analysis of amino acid mutations on fusion inhibition for the inhibitor SCH-C. Y108A, I198M and L33A which were shown to have a significant Mutation Mean IC50 SEM N contribution to inhibitor binding, and A29G and Y251F which had a minor contribution (Billick et al., 2004; Dragic et al., 2000; Castonguay Wild type 4.7 0.92 72 K26A 7.2 0.75 17 et al., 2003; Kondru et al., 2008; Maeda et al., 2006; Nishikawa et al., A29G 7.8 1.56 6 2005; Seibert et al., 2006; Stupple et al., 2011; Tsamis et al., 2003). The R31A 4.9 0.65 9 effects of these mutations on receptor/compound interaction on the L33A 131.3 8.93 41 small molecule antagonists shown in Fig.